Klasifikasi Citra Kain Songke Manggarai Timur Berdasarkan Tekstur Dan Warna Menggunakan Metode K-Nearest Neighbors (KNN)

RISTIE, Eugenia Salmaliani Alvira (2024) Klasifikasi Citra Kain Songke Manggarai Timur Berdasarkan Tekstur Dan Warna Menggunakan Metode K-Nearest Neighbors (KNN). Undergraduate thesis, Universitas Katolik Widya Mandira Kupang.

	Text ABSTRAK.pdf Download (763kB)
	Text BAB I.pdf Download (153kB)
	Text BAB II.pdf Restricted to Repository staff only Download (259kB)
	Text BAB III.pdf Restricted to Repository staff only Download (191kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (509kB)
	Text BAB V.pdf Restricted to Repository staff only Download (98kB)
	Text BAB VI.pdf Download (12kB)
	Text DAFTAR PUSTAKA DAN SUKET BEBAS PLAGIAT.pdf Download (803kB)

Abstract

This research aims to develop a classification model of East Manggarai songke cloth images based on texture and color motifs using the Machine Learning K-Nearest Neighbors (KNN) method. This method is employed to identify and distinguish motifs such as lines, mata manuk (chicken eye), nyala (flame), ranggong (spider), wela ngkaweng, and wela runu. The dataset consists of 450 annotated images with corresponding motifs. Testing was conducted using 20-fold cross- validation to ensure objective evaluation and achieve reliable Accuracy. The results show an overall Accuracy of 97.78%, with high precision, recall, and F1-score for each motif class. However, there were some minor misclassifications among similar motifs, indicating potential areas for model improvement. This study contributes to the application of Machine Learning in understanding and preserving the cultural richness of East Manggarai traditional cloth.

Actions (login required)

View Item